The goal of the proposed research is to develop and validate an integrated method for non-invasive dynamic imaging of human brain function. In the proposed method, structural MRI is used to identify possible MEG/EEG generating dipoles as lying within the cortex and oriented perpendicular to its surface. A linear approach is used to estimate the time-course of activation at every point on the cortical surface, based on observed EM field patterns. The solution is biased toward those brain areas found to be activated using fMRI acquired in the same subject in the same task. Preliminary studies suggest that an anatomical and functional data approach can provide accurate spatiotemporal maps of cerebral activity. Potential sources of error are "missing" sources and head model mis-specification, where the missing sources arise possibly due to paradigm differences or data artifacts. In Aim 1, the applicants proposed to develop acquisition techniques that permit the identical task paradigm to be used in all experiments, and proposed improvements in their fMRI measurements to reduce inherent magnetic susceptibility artifacts. Another potential source of error is caused by inaccuracies in the forward solution due to incorrect estimates of tissue conductivities. In Aim 2, the applicants proposed to develop and test a novel method for non- invasively measuring the tissue conductivity of the brain using MR diffusion imaging. Although simulation studies suggest how to best combine data, the accuracy of the spatio-temporal maps is difficult to determine in humans. In Aim 3, the applicants proposed to directly validate their non-invasive estimates via direct comparison with intracranial electrode measurements in patients.